Retrieving ice cloud properties by using a fast infrared radiative transfer model

A new fast infrared radiative transfer (RT) model based on pre-computed look-up tables (LUTs) including the LUTs for emissivity function and cloud effective temperature is proposed. This model can be applied to the simulation of upward radiance (or brightness temperature) at 8.5, 11.0 and 12.0 μm at the top of the atmosphere (TOA) under cloudy-sky conditions. Optical depths of Atmospheric layers resulting from gaseous absorption are derived from the correlated-K distribution (CKD) method. The cloud reflection and transmission functions are computed from the discrete ordinates radiative transfer model (DISORT). In addition to the LUTs of reflection and transmission functions of cloud in traditional RT models, the LUTs of emissivity and effective temperature are also included to improve the accuracy. Generally speaking, for an atmosphere containing a single ice cloud layer with small optical thickness (i.e., less than 5.0), the brightness temperature differences (BTDs) between the fast model and DISORT results are approximately less than 0.1K, whereas the BTDs are less than 0.02K when the ice cloud optical thickness is larger than 5.0. Moreover, with the fast RT model, cloud optical and microphysical properties of ice clouds are retrieved from MODIS and CALIPSO observations and the MERRA reanalysis data. The present retrievals are compared with the MODIS operational cloud products (MYD06).